用风云二号C星进行热点监测的快速算法  

A Quick Algorithm of Hot Spot Monitoring Using FY-2C Satellite Data

作　　者：杨晓峰[1] 郑有飞[1] 温兴平[2] 张育萍 

机构地区：[1]南京信息工程大学环境科学与工程学院,南京210044 [2]中国地质大学数学地质遥感地质研究所,武汉430074 [3]江苏省张家港市气象局,张家港215600

出　　处：《遥感信息》2008年第1期43-46,共4页Remote Sensing Information

摘　　要：用卫星影像资料监测热点已经有较长的历史。静止卫星由于影像采集时间间隔短,能显著提高热点监测的及时性,但由于其星下点分辨率低,影像覆盖范围广,监测热点较为困难。文中提出了一种利用风云二号C星影像资料进行热点监测的快速算法,首先对影像资料进行辐射定标并生成中红外与长红外通道的差值影像图,而后将中红外影像图与差值影像图分割为20×20像元的小块,分别计算各小块的最大值、平均值及标准差,认为两幅影像相对应的像元块内最大值均大于平均值3倍标准差即认为该块内可能含有热点;然后结合地表真实分布情况对含有热点的小块进行综合识别处理;最后将含有热点的局部影像进行几何校正投影输出。通过实例资料进行检验,认为这种方法计算简单、快速,可极大提高热点监测的准确性与时效性。It has a long history to monitor the surface hot spot using remote sensing technology. Stationary satellite can monitor the surface hot spot timely for its short acquiring interval. However, due to its lower nadir resolution and larger covering areas, it is hard to monitor the surface hot spot exactly. In this paper, a quick algorithm is put forward to monitor the surface hot spot using FY-2C satellite. Firstly, the remote sensing image should be calibrated to brightness temperature and an image is produced by subtracting the long wave infrared channel data from the medium wave infrared data. Secondly, the medium wave infrared and the subtracting image should be divided into 20 by 20 pixels sub-images respectively. Then the sub-images of corresponding images whose maximum values are three times greater than mean value may be contained the hot spot. Then, the hot spot areas are distinguished from these sub-images by comparing with the real surface data. Finally, the remote sensing image is geometrically transformed and printed. This algorithm was tested by the data from FY-2C satellite when there was a forest fire along the border of China and Mongolia. It is concluded that this is a simple and quick algorithm which can improve the precision and time of monitoring the hot spot greatly.

关 键 词：风云二号C星 静止卫星 热点监测 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]